MNIST Classification using Neural ODEs

To understand Neural ODEs, users should look up these lecture notes. We recommend users to directly use DiffEqFlux.jl, instead of implementing Neural ODEs from scratch.

Package Imports

using Lux, ComponentArrays, SciMLSensitivity, LuxCUDA, Optimisers, OrdinaryDiffEq, Random,
      Statistics, Zygote, OneHotArrays, InteractiveUtils, Printf
using MLDatasets: MNIST
using MLUtils: DataLoader, splitobs

CUDA.allowscalar(false)

Loading MNIST

function loadmnist(batchsize, train_split)
    # Load MNIST: Only 1500 for demonstration purposes
    N = 1500
    dataset = MNIST(; split=:train)
    imgs = dataset.features[:, :, 1:N]
    labels_raw = dataset.targets[1:N]

    # Process images into (H,W,C,BS) batches
    x_data = Float32.(reshape(imgs, size(imgs, 1), size(imgs, 2), 1, size(imgs, 3)))
    y_data = onehotbatch(labels_raw, 0:9)
    (x_train, y_train), (x_test, y_test) = splitobs((x_data, y_data); at=train_split)

    return (
        # Use DataLoader to automatically minibatch and shuffle the data
        DataLoader(collect.((x_train, y_train)); batchsize, shuffle=true),
        # Don't shuffle the test data
        DataLoader(collect.((x_test, y_test)); batchsize, shuffle=false))
end

loadmnist (generic function with 1 method)

Define the Neural ODE Layer

First we will use the @compact macro to define the Neural ODE Layer.

function NeuralODECompact(
        model::Lux.AbstractLuxLayer; solver=Tsit5(), tspan=(0.0f0, 1.0f0), kwargs...)
    return @compact(; model, solver, tspan, kwargs...) do x, p
        dudt(u, p, t) = vec(model(reshape(u, size(x)), p))
        # Note the `p.model` here
        prob = ODEProblem(ODEFunction{false}(dudt), vec(x), tspan, p.model)
        @return solve(prob, solver; kwargs...)
    end
end

NeuralODECompact (generic function with 1 method)

We recommend using the compact macro for creating custom layers. The below implementation exists mostly for historical reasons when @compact was not part of the stable API. Also, it helps users understand how the layer interface of Lux works.

The NeuralODE is a ContainerLayer, which stores a model. The parameters and states of the NeuralODE are same as those of the underlying model.

struct NeuralODE{M <: Lux.AbstractLuxLayer, So, T, K} <: Lux.AbstractLuxWrapperLayer{:model}
    model::M
    solver::So
    tspan::T
    kwargs::K
end

function NeuralODE(
        model::Lux.AbstractLuxLayer; solver=Tsit5(), tspan=(0.0f0, 1.0f0), kwargs...)
    return NeuralODE(model, solver, tspan, kwargs)
end

Main.var"##225".NeuralODE

OrdinaryDiffEq.jl can deal with non-Vector Inputs! However, certain discrete sensitivities like ReverseDiffAdjoint can't handle non-Vector inputs. Hence, we need to convert the input and output of the ODE solver to a Vector.

function (n::NeuralODE)(x, ps, st)
    function dudt(u, p, t)
        u_, st = n.model(reshape(u, size(x)), p, st)
        return vec(u_)
    end
    prob = ODEProblem{false}(ODEFunction{false}(dudt), vec(x), n.tspan, ps)
    return solve(prob, n.solver; n.kwargs...), st
end

@views diffeqsol_to_array(l::Int, x::ODESolution) = reshape(last(x.u), (l, :))
@views diffeqsol_to_array(l::Int, x::AbstractMatrix) = reshape(x[:, end], (l, :))

diffeqsol_to_array (generic function with 2 methods)

Create and Initialize the Neural ODE Layer

function create_model(model_fn=NeuralODE; dev=gpu_device(), use_named_tuple::Bool=false,
        sensealg=InterpolatingAdjoint(; autojacvec=ZygoteVJP()))
    # Construct the Neural ODE Model
    model = Chain(FlattenLayer(),
        Dense(784 => 20, tanh),
        model_fn(Chain(Dense(20 => 10, tanh), Dense(10 => 10, tanh), Dense(10 => 20, tanh));
            save_everystep=false, reltol=1.0f-3,
            abstol=1.0f-3, save_start=false, sensealg),
        Base.Fix1(diffeqsol_to_array, 20),
        Dense(20 => 10))

    rng = Random.default_rng()
    Random.seed!(rng, 0)

    ps, st = Lux.setup(rng, model)
    ps = (use_named_tuple ? ps : ComponentArray(ps)) |> dev
    st = st |> dev

    return model, ps, st
end

create_model (generic function with 2 methods)

Define Utility Functions

const logitcrossentropy = CrossEntropyLoss(; logits=Val(true))

function accuracy(model, ps, st, dataloader)
    total_correct, total = 0, 0
    st = Lux.testmode(st)
    for (x, y) in dataloader
        target_class = onecold(y)
        predicted_class = onecold(first(model(x, ps, st)))
        total_correct += sum(target_class .== predicted_class)
        total += length(target_class)
    end
    return total_correct / total
end

accuracy (generic function with 1 method)

Training

function train(model_function; cpu::Bool=false, kwargs...)
    dev = cpu ? cpu_device() : gpu_device()
    model, ps, st = create_model(model_function; dev, kwargs...)

    # Training
    train_dataloader, test_dataloader = loadmnist(128, 0.9) |> dev

    tstate = Training.TrainState(model, ps, st, Adam(0.001f0))

    ### Lets train the model
    nepochs = 9
    for epoch in 1:nepochs
        stime = time()
        for (x, y) in train_dataloader
            _, _, _, tstate = Training.single_train_step!(
                AutoZygote(), logitcrossentropy, (x, y), tstate)
        end
        ttime = time() - stime

        tr_acc = accuracy(model, tstate.parameters, tstate.states, train_dataloader)
        te_acc = accuracy(model, tstate.parameters, tstate.states, test_dataloader)
        @printf "[%d/%d] \t Time %.2fs \t Training Accuracy: %.5f%% \t Test \
                 Accuracy: %.5f%%\n" epoch nepochs ttime tr_acc te_acc
    end
end

train(NeuralODECompact)

[1/9] 	 Time 123.36s 	 Training Accuracy: 0.37481% 	 Test Accuracy: 0.40000%
[2/9] 	 Time 0.53s 	 Training Accuracy: 0.57556% 	 Test Accuracy: 0.54667%
[3/9] 	 Time 0.53s 	 Training Accuracy: 0.68296% 	 Test Accuracy: 0.70000%
[4/9] 	 Time 0.53s 	 Training Accuracy: 0.74148% 	 Test Accuracy: 0.74667%
[5/9] 	 Time 0.53s 	 Training Accuracy: 0.76444% 	 Test Accuracy: 0.78000%
[6/9] 	 Time 0.52s 	 Training Accuracy: 0.80000% 	 Test Accuracy: 0.78667%
[7/9] 	 Time 0.53s 	 Training Accuracy: 0.80889% 	 Test Accuracy: 0.80667%
[8/9] 	 Time 0.52s 	 Training Accuracy: 0.83185% 	 Test Accuracy: 0.80667%
[9/9] 	 Time 0.69s 	 Training Accuracy: 0.84815% 	 Test Accuracy: 0.81333%

train(NeuralODE)

[1/9] 	 Time 34.02s 	 Training Accuracy: 0.37481% 	 Test Accuracy: 0.40000%
[2/9] 	 Time 0.54s 	 Training Accuracy: 0.57926% 	 Test Accuracy: 0.58000%
[3/9] 	 Time 0.55s 	 Training Accuracy: 0.68519% 	 Test Accuracy: 0.69333%
[4/9] 	 Time 0.53s 	 Training Accuracy: 0.73333% 	 Test Accuracy: 0.73333%
[5/9] 	 Time 0.53s 	 Training Accuracy: 0.76074% 	 Test Accuracy: 0.77333%
[6/9] 	 Time 0.75s 	 Training Accuracy: 0.79630% 	 Test Accuracy: 0.80000%
[7/9] 	 Time 0.53s 	 Training Accuracy: 0.81037% 	 Test Accuracy: 0.80667%
[8/9] 	 Time 0.53s 	 Training Accuracy: 0.84222% 	 Test Accuracy: 0.80000%
[9/9] 	 Time 0.74s 	 Training Accuracy: 0.85778% 	 Test Accuracy: 0.82667%

We can also change the sensealg and train the model! GaussAdjoint allows you to use any arbitrary parameter structure and not just a flat vector (ComponentArray).

train(NeuralODE; sensealg=GaussAdjoint(; autojacvec=ZygoteVJP()), use_named_tuple=true)

[1/9] 	 Time 39.10s 	 Training Accuracy: 0.37481% 	 Test Accuracy: 0.40000%
[2/9] 	 Time 0.50s 	 Training Accuracy: 0.57556% 	 Test Accuracy: 0.54000%
[3/9] 	 Time 0.52s 	 Training Accuracy: 0.69852% 	 Test Accuracy: 0.69333%
[4/9] 	 Time 0.69s 	 Training Accuracy: 0.72519% 	 Test Accuracy: 0.74000%
[5/9] 	 Time 0.50s 	 Training Accuracy: 0.75333% 	 Test Accuracy: 0.76000%
[6/9] 	 Time 0.67s 	 Training Accuracy: 0.78889% 	 Test Accuracy: 0.79333%
[7/9] 	 Time 0.50s 	 Training Accuracy: 0.81037% 	 Test Accuracy: 0.80000%
[8/9] 	 Time 0.50s 	 Training Accuracy: 0.83778% 	 Test Accuracy: 0.81333%
[9/9] 	 Time 0.69s 	 Training Accuracy: 0.85259% 	 Test Accuracy: 0.82667%

But remember some AD backends like ReverseDiff is not GPU compatible. For a model this size, you will notice that training time is significantly lower for training on CPU than on GPU.

train(NeuralODE; sensealg=InterpolatingAdjoint(; autojacvec=ReverseDiffVJP()), cpu=true)

[1/9] 	 Time 108.21s 	 Training Accuracy: 0.37481% 	 Test Accuracy: 0.40000%
[2/9] 	 Time 22.29s 	 Training Accuracy: 0.58741% 	 Test Accuracy: 0.56667%
[3/9] 	 Time 24.51s 	 Training Accuracy: 0.69926% 	 Test Accuracy: 0.71333%
[4/9] 	 Time 23.52s 	 Training Accuracy: 0.72815% 	 Test Accuracy: 0.74000%
[5/9] 	 Time 24.84s 	 Training Accuracy: 0.76370% 	 Test Accuracy: 0.78667%
[6/9] 	 Time 20.17s 	 Training Accuracy: 0.79037% 	 Test Accuracy: 0.80667%
[7/9] 	 Time 26.08s 	 Training Accuracy: 0.81630% 	 Test Accuracy: 0.80667%
[8/9] 	 Time 25.52s 	 Training Accuracy: 0.83333% 	 Test Accuracy: 0.80000%
[9/9] 	 Time 23.65s 	 Training Accuracy: 0.85407% 	 Test Accuracy: 0.82000%

For completeness, let's also test out discrete sensitivities!

train(NeuralODE; sensealg=ReverseDiffAdjoint(), cpu=true)

[1/9] 	 Time 47.82s 	 Training Accuracy: 0.37481% 	 Test Accuracy: 0.40000%
[2/9] 	 Time 24.01s 	 Training Accuracy: 0.58667% 	 Test Accuracy: 0.57333%
[3/9] 	 Time 21.54s 	 Training Accuracy: 0.69704% 	 Test Accuracy: 0.71333%
[4/9] 	 Time 21.08s 	 Training Accuracy: 0.72741% 	 Test Accuracy: 0.74000%
[5/9] 	 Time 28.25s 	 Training Accuracy: 0.76148% 	 Test Accuracy: 0.78667%
[6/9] 	 Time 28.00s 	 Training Accuracy: 0.79037% 	 Test Accuracy: 0.80667%
[7/9] 	 Time 28.15s 	 Training Accuracy: 0.81556% 	 Test Accuracy: 0.80667%
[8/9] 	 Time 29.23s 	 Training Accuracy: 0.83407% 	 Test Accuracy: 0.80000%
[9/9] 	 Time 28.14s 	 Training Accuracy: 0.85259% 	 Test Accuracy: 0.81333%

Alternate Implementation using Stateful Layer

Starting v0.5.5, Lux provides a StatefulLuxLayer which can be used to avoid the Boxing of st. Using the @compact API avoids this problem entirely.

struct StatefulNeuralODE{M <: Lux.AbstractLuxLayer, So, T, K} <:
       Lux.AbstractLuxWrapperLayer{:model}
    model::M
    solver::So
    tspan::T
    kwargs::K
end

function StatefulNeuralODE(
        model::Lux.AbstractLuxLayer; solver=Tsit5(), tspan=(0.0f0, 1.0f0), kwargs...)
    return StatefulNeuralODE(model, solver, tspan, kwargs)
end

function (n::StatefulNeuralODE)(x, ps, st)
    st_model = StatefulLuxLayer{true}(n.model, ps, st)
    dudt(u, p, t) = st_model(u, p)
    prob = ODEProblem{false}(ODEFunction{false}(dudt), x, n.tspan, ps)
    return solve(prob, n.solver; n.kwargs...), st_model.st
end

Train the new Stateful Neural ODE

train(StatefulNeuralODE)

[1/9] 	 Time 36.88s 	 Training Accuracy: 0.37481% 	 Test Accuracy: 0.40000%
[2/9] 	 Time 0.49s 	 Training Accuracy: 0.57926% 	 Test Accuracy: 0.58000%
[3/9] 	 Time 0.51s 	 Training Accuracy: 0.67704% 	 Test Accuracy: 0.68667%
[4/9] 	 Time 0.73s 	 Training Accuracy: 0.73259% 	 Test Accuracy: 0.73333%
[5/9] 	 Time 0.50s 	 Training Accuracy: 0.76593% 	 Test Accuracy: 0.76667%
[6/9] 	 Time 0.49s 	 Training Accuracy: 0.79852% 	 Test Accuracy: 0.78667%
[7/9] 	 Time 0.52s 	 Training Accuracy: 0.80593% 	 Test Accuracy: 0.80000%
[8/9] 	 Time 0.50s 	 Training Accuracy: 0.83556% 	 Test Accuracy: 0.80667%
[9/9] 	 Time 0.49s 	 Training Accuracy: 0.83704% 	 Test Accuracy: 0.82667%

We might not see a significant difference in the training time, but let us investigate the type stabilities of the layers.

Type Stability

model, ps, st = create_model(NeuralODE)

model_stateful, ps_stateful, st_stateful = create_model(StatefulNeuralODE)

x = gpu_device()(ones(Float32, 28, 28, 1, 3));

NeuralODE is not type stable due to the boxing of st

@code_warntype model(x, ps, st)

MethodInstance for (::Lux.Chain{@NamedTuple{layer_1::Lux.FlattenLayer{Nothing}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Main.var"##225".NeuralODE{Lux.Chain{@NamedTuple{layer_1::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}, OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}, Tuple{Float32, Float32}, Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}, layer_4::Lux.WrappedFunction{Base.Fix1{typeof(Main.var"##225".diffeqsol_to_array), Int64}}, layer_5::Lux.Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing})(::CUDA.CuArray{Float32, 4, CUDA.DeviceMemory}, ::ComponentArrays.ComponentVector{Float32, CUDA.CuArray{Float32, 1, CUDA.DeviceMemory}, Tuple{ComponentArrays.Axis{(layer_1 = 1:0, layer_2 = ViewAxis(1:15700, Axis(weight = ViewAxis(1:15680, ShapedAxis((20, 784))), bias = 15681:15700)), layer_3 = ViewAxis(15701:16240, Axis(layer_1 = ViewAxis(1:210, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)), layer_2 = ViewAxis(211:320, Axis(weight = ViewAxis(1:100, ShapedAxis((10, 10))), bias = 101:110)), layer_3 = ViewAxis(321:540, Axis(weight = ViewAxis(1:200, ShapedAxis((20, 10))), bias = 201:220)))), layer_4 = 16241:16240, layer_5 = ViewAxis(16241:16450, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)))}}}, ::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{}}, layer_4::@NamedTuple{}, layer_5::@NamedTuple{}})
  from (c::Lux.Chain)(x, ps, st::NamedTuple) @ Lux /var/lib/buildkite-agent/builds/gpuci-4/julialang/lux-dot-jl/src/layers/containers.jl:480
Arguments
  c::Lux.Chain{@NamedTuple{layer_1::Lux.FlattenLayer{Nothing}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Main.var"##225".NeuralODE{Lux.Chain{@NamedTuple{layer_1::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}, OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}, Tuple{Float32, Float32}, Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}, layer_4::Lux.WrappedFunction{Base.Fix1{typeof(Main.var"##225".diffeqsol_to_array), Int64}}, layer_5::Lux.Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}
  x::CUDA.CuArray{Float32, 4, CUDA.DeviceMemory}
  ps::ComponentArrays.ComponentVector{Float32, CUDA.CuArray{Float32, 1, CUDA.DeviceMemory}, Tuple{ComponentArrays.Axis{(layer_1 = 1:0, layer_2 = ViewAxis(1:15700, Axis(weight = ViewAxis(1:15680, ShapedAxis((20, 784))), bias = 15681:15700)), layer_3 = ViewAxis(15701:16240, Axis(layer_1 = ViewAxis(1:210, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)), layer_2 = ViewAxis(211:320, Axis(weight = ViewAxis(1:100, ShapedAxis((10, 10))), bias = 101:110)), layer_3 = ViewAxis(321:540, Axis(weight = ViewAxis(1:200, ShapedAxis((20, 10))), bias = 201:220)))), layer_4 = 16241:16240, layer_5 = ViewAxis(16241:16450, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)))}}}
  st::Core.Const((layer_1 = NamedTuple(), layer_2 = NamedTuple(), layer_3 = (layer_1 = NamedTuple(), layer_2 = NamedTuple(), layer_3 = NamedTuple()), layer_4 = NamedTuple(), layer_5 = NamedTuple()))
Body::TUPLE{CUDA.CUARRAY{FLOAT32, 2, CUDA.DEVICEMEMORY}, NAMEDTUPLE{(:LAYER_1, :LAYER_2, :LAYER_3, :LAYER_4, :LAYER_5), <:TUPLE{@NAMEDTUPLE{}, @NAMEDTUPLE{}, ANY, @NAMEDTUPLE{}, @NAMEDTUPLE{}}}}
1 ─ %1 = Base.getproperty(c, :layers)::@NamedTuple{layer_1::Lux.FlattenLayer{Nothing}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Main.var"##225".NeuralODE{Lux.Chain{@NamedTuple{layer_1::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}, OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}, Tuple{Float32, Float32}, Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}, layer_4::Lux.WrappedFunction{Base.Fix1{typeof(Main.var"##225".diffeqsol_to_array), Int64}}, layer_5::Lux.Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}
│   %2 = Lux.applychain(%1, x, ps, st)::TUPLE{CUDA.CUARRAY{FLOAT32, 2, CUDA.DEVICEMEMORY}, NAMEDTUPLE{(:LAYER_1, :LAYER_2, :LAYER_3, :LAYER_4, :LAYER_5), <:TUPLE{@NAMEDTUPLE{}, @NAMEDTUPLE{}, ANY, @NAMEDTUPLE{}, @NAMEDTUPLE{}}}}
└──      return %2

We avoid the problem entirely by using StatefulNeuralODE

@code_warntype model_stateful(x, ps_stateful, st_stateful)

MethodInstance for (::Lux.Chain{@NamedTuple{layer_1::Lux.FlattenLayer{Nothing}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Main.var"##225".StatefulNeuralODE{Lux.Chain{@NamedTuple{layer_1::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}, OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}, Tuple{Float32, Float32}, Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}, layer_4::Lux.WrappedFunction{Base.Fix1{typeof(Main.var"##225".diffeqsol_to_array), Int64}}, layer_5::Lux.Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing})(::CUDA.CuArray{Float32, 4, CUDA.DeviceMemory}, ::ComponentArrays.ComponentVector{Float32, CUDA.CuArray{Float32, 1, CUDA.DeviceMemory}, Tuple{ComponentArrays.Axis{(layer_1 = 1:0, layer_2 = ViewAxis(1:15700, Axis(weight = ViewAxis(1:15680, ShapedAxis((20, 784))), bias = 15681:15700)), layer_3 = ViewAxis(15701:16240, Axis(layer_1 = ViewAxis(1:210, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)), layer_2 = ViewAxis(211:320, Axis(weight = ViewAxis(1:100, ShapedAxis((10, 10))), bias = 101:110)), layer_3 = ViewAxis(321:540, Axis(weight = ViewAxis(1:200, ShapedAxis((20, 10))), bias = 201:220)))), layer_4 = 16241:16240, layer_5 = ViewAxis(16241:16450, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)))}}}, ::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{}}, layer_4::@NamedTuple{}, layer_5::@NamedTuple{}})
  from (c::Lux.Chain)(x, ps, st::NamedTuple) @ Lux /var/lib/buildkite-agent/builds/gpuci-4/julialang/lux-dot-jl/src/layers/containers.jl:480
Arguments
  c::Lux.Chain{@NamedTuple{layer_1::Lux.FlattenLayer{Nothing}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Main.var"##225".StatefulNeuralODE{Lux.Chain{@NamedTuple{layer_1::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}, OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}, Tuple{Float32, Float32}, Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}, layer_4::Lux.WrappedFunction{Base.Fix1{typeof(Main.var"##225".diffeqsol_to_array), Int64}}, layer_5::Lux.Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}
  x::CUDA.CuArray{Float32, 4, CUDA.DeviceMemory}
  ps::ComponentArrays.ComponentVector{Float32, CUDA.CuArray{Float32, 1, CUDA.DeviceMemory}, Tuple{ComponentArrays.Axis{(layer_1 = 1:0, layer_2 = ViewAxis(1:15700, Axis(weight = ViewAxis(1:15680, ShapedAxis((20, 784))), bias = 15681:15700)), layer_3 = ViewAxis(15701:16240, Axis(layer_1 = ViewAxis(1:210, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)), layer_2 = ViewAxis(211:320, Axis(weight = ViewAxis(1:100, ShapedAxis((10, 10))), bias = 101:110)), layer_3 = ViewAxis(321:540, Axis(weight = ViewAxis(1:200, ShapedAxis((20, 10))), bias = 201:220)))), layer_4 = 16241:16240, layer_5 = ViewAxis(16241:16450, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)))}}}
  st::Core.Const((layer_1 = NamedTuple(), layer_2 = NamedTuple(), layer_3 = (layer_1 = NamedTuple(), layer_2 = NamedTuple(), layer_3 = NamedTuple()), layer_4 = NamedTuple(), layer_5 = NamedTuple()))
Body::Tuple{CUDA.CuArray{Float32, 2, CUDA.DeviceMemory}, @NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{}}, layer_4::@NamedTuple{}, layer_5::@NamedTuple{}}}
1 ─ %1 = Base.getproperty(c, :layers)::@NamedTuple{layer_1::Lux.FlattenLayer{Nothing}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Main.var"##225".StatefulNeuralODE{Lux.Chain{@NamedTuple{layer_1::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}, OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}, Tuple{Float32, Float32}, Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}, layer_4::Lux.WrappedFunction{Base.Fix1{typeof(Main.var"##225".diffeqsol_to_array), Int64}}, layer_5::Lux.Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}
│   %2 = Lux.applychain(%1, x, ps, st)::Tuple{CUDA.CuArray{Float32, 2, CUDA.DeviceMemory}, @NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{}}, layer_4::@NamedTuple{}, layer_5::@NamedTuple{}}}
└──      return %2

Note, that we still recommend using this layer internally and not exposing this as the default API to the users.

Finally checking the compact model

model_compact, ps_compact, st_compact = create_model(NeuralODECompact)

@code_warntype model_compact(x, ps_compact, st_compact)

MethodInstance for (::Lux.Chain{@NamedTuple{layer_1::Lux.FlattenLayer{Nothing}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.CompactLuxLayer{:₋₋₋no_special_dispatch₋₋₋, Main.var"##225".var"#2#3", Nothing, @NamedTuple{model::Lux.Chain{@NamedTuple{layer_1::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}}, Lux.CompactMacroImpl.ValueStorage{@NamedTuple{}, @NamedTuple{solver::Returns{OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}}, tspan::Returns{Tuple{Float32, Float32}}}}, Tuple{Tuple{Symbol}, Tuple{Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}}}, layer_4::Lux.WrappedFunction{Base.Fix1{typeof(Main.var"##225".diffeqsol_to_array), Int64}}, layer_5::Lux.Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing})(::CUDA.CuArray{Float32, 4, CUDA.DeviceMemory}, ::ComponentArrays.ComponentVector{Float32, CUDA.CuArray{Float32, 1, CUDA.DeviceMemory}, Tuple{ComponentArrays.Axis{(layer_1 = 1:0, layer_2 = ViewAxis(1:15700, Axis(weight = ViewAxis(1:15680, ShapedAxis((20, 784))), bias = 15681:15700)), layer_3 = ViewAxis(15701:16240, Axis(model = ViewAxis(1:540, Axis(layer_1 = ViewAxis(1:210, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)), layer_2 = ViewAxis(211:320, Axis(weight = ViewAxis(1:100, ShapedAxis((10, 10))), bias = 101:110)), layer_3 = ViewAxis(321:540, Axis(weight = ViewAxis(1:200, ShapedAxis((20, 10))), bias = 201:220)))),)), layer_4 = 16241:16240, layer_5 = ViewAxis(16241:16450, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)))}}}, ::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{model::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{}}, solver::OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}, tspan::Tuple{Float32, Float32}, ₋₋₋kwargs₋₋₋::@NamedTuple{kwargs::Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}}, layer_4::@NamedTuple{}, layer_5::@NamedTuple{}})
  from (c::Lux.Chain)(x, ps, st::NamedTuple) @ Lux /var/lib/buildkite-agent/builds/gpuci-4/julialang/lux-dot-jl/src/layers/containers.jl:480
Arguments
  c::Lux.Chain{@NamedTuple{layer_1::Lux.FlattenLayer{Nothing}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.CompactLuxLayer{:₋₋₋no_special_dispatch₋₋₋, Main.var"##225".var"#2#3", Nothing, @NamedTuple{model::Lux.Chain{@NamedTuple{layer_1::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}}, Lux.CompactMacroImpl.ValueStorage{@NamedTuple{}, @NamedTuple{solver::Returns{OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}}, tspan::Returns{Tuple{Float32, Float32}}}}, Tuple{Tuple{Symbol}, Tuple{Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}}}, layer_4::Lux.WrappedFunction{Base.Fix1{typeof(Main.var"##225".diffeqsol_to_array), Int64}}, layer_5::Lux.Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}
  x::CUDA.CuArray{Float32, 4, CUDA.DeviceMemory}
  ps::ComponentArrays.ComponentVector{Float32, CUDA.CuArray{Float32, 1, CUDA.DeviceMemory}, Tuple{ComponentArrays.Axis{(layer_1 = 1:0, layer_2 = ViewAxis(1:15700, Axis(weight = ViewAxis(1:15680, ShapedAxis((20, 784))), bias = 15681:15700)), layer_3 = ViewAxis(15701:16240, Axis(model = ViewAxis(1:540, Axis(layer_1 = ViewAxis(1:210, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)), layer_2 = ViewAxis(211:320, Axis(weight = ViewAxis(1:100, ShapedAxis((10, 10))), bias = 101:110)), layer_3 = ViewAxis(321:540, Axis(weight = ViewAxis(1:200, ShapedAxis((20, 10))), bias = 201:220)))),)), layer_4 = 16241:16240, layer_5 = ViewAxis(16241:16450, Axis(weight = ViewAxis(1:200, ShapedAxis((10, 20))), bias = 201:210)))}}}
  st::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{model::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{}}, solver::OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}, tspan::Tuple{Float32, Float32}, ₋₋₋kwargs₋₋₋::@NamedTuple{kwargs::Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}}, layer_4::@NamedTuple{}, layer_5::@NamedTuple{}}
Body::Tuple{CUDA.CuArray{Float32, 2, CUDA.DeviceMemory}, @NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{model::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{}}, solver::OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}, tspan::Tuple{Float32, Float32}, ₋₋₋kwargs₋₋₋::@NamedTuple{kwargs::Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}}, layer_4::@NamedTuple{}, layer_5::@NamedTuple{}}}
1 ─ %1 = Base.getproperty(c, :layers)::@NamedTuple{layer_1::Lux.FlattenLayer{Nothing}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.CompactLuxLayer{:₋₋₋no_special_dispatch₋₋₋, Main.var"##225".var"#2#3", Nothing, @NamedTuple{model::Lux.Chain{@NamedTuple{layer_1::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_3::Lux.Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}}, Lux.CompactMacroImpl.ValueStorage{@NamedTuple{}, @NamedTuple{solver::Returns{OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}}, tspan::Returns{Tuple{Float32, Float32}}}}, Tuple{Tuple{Symbol}, Tuple{Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}}}, layer_4::Lux.WrappedFunction{Base.Fix1{typeof(Main.var"##225".diffeqsol_to_array), Int64}}, layer_5::Lux.Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}
│   %2 = Lux.applychain(%1, x, ps, st)::Tuple{CUDA.CuArray{Float32, 2, CUDA.DeviceMemory}, @NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{model::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}, layer_3::@NamedTuple{}}, solver::OrdinaryDiffEqTsit5.Tsit5{typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!), Static.False}, tspan::Tuple{Float32, Float32}, ₋₋₋kwargs₋₋₋::@NamedTuple{kwargs::Base.Pairs{Symbol, Any, NTuple{5, Symbol}, @NamedTuple{save_everystep::Bool, reltol::Float32, abstol::Float32, save_start::Bool, sensealg::SciMLSensitivity.InterpolatingAdjoint{0, true, Val{:central}, SciMLSensitivity.ZygoteVJP}}}}}, layer_4::@NamedTuple{}, layer_5::@NamedTuple{}}}
└──      return %2

Appendix

using InteractiveUtils
InteractiveUtils.versioninfo()

if @isdefined(MLDataDevices)
    if @isdefined(CUDA) && MLDataDevices.functional(CUDADevice)
        println()
        CUDA.versioninfo()
    end

    if @isdefined(AMDGPU) && MLDataDevices.functional(AMDGPUDevice)
        println()
        AMDGPU.versioninfo()
    end
end

Julia Version 1.10.5
Commit 6f3fdf7b362 (2024-08-27 14:19 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 48 × AMD EPYC 7402 24-Core Processor
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-15.0.7 (ORCJIT, znver2)
Threads: 48 default, 0 interactive, 24 GC (on 2 virtual cores)
Environment:
  JULIA_CPU_THREADS = 2
  JULIA_DEPOT_PATH = /root/.cache/julia-buildkite-plugin/depots/01872db4-8c79-43af-ab7d-12abac4f24f6
  LD_LIBRARY_PATH = /usr/local/nvidia/lib:/usr/local/nvidia/lib64
  JULIA_PKG_SERVER = 
  JULIA_NUM_THREADS = 48
  JULIA_CUDA_HARD_MEMORY_LIMIT = 100%
  JULIA_PKG_PRECOMPILE_AUTO = 0
  JULIA_DEBUG = Literate

CUDA runtime 12.5, artifact installation
CUDA driver 12.5
NVIDIA driver 555.42.6

CUDA libraries: 
- CUBLAS: 12.5.3
- CURAND: 10.3.6
- CUFFT: 11.2.3
- CUSOLVER: 11.6.3
- CUSPARSE: 12.5.1
- CUPTI: 2024.2.1 (API 23.0.0)
- NVML: 12.0.0+555.42.6

Julia packages: 
- CUDA: 5.4.3
- CUDA_Driver_jll: 0.9.2+0
- CUDA_Runtime_jll: 0.14.1+0

Toolchain:
- Julia: 1.10.5
- LLVM: 15.0.7

Environment:
- JULIA_CUDA_HARD_MEMORY_LIMIT: 100%

1 device:
  0: NVIDIA A100-PCIE-40GB MIG 1g.5gb (sm_80, 3.951 GiB / 4.750 GiB available)

---

This page was generated using Literate.jl.